ECI Symposium Series, Volume RP5: Proceedings of 7th International Conference on Heat Exchanger Fouling and Cleaning Challenges

and Opportunities, Editors Hans Müller-Steinhagen, M. Reza Malayeri, and A. Paul Watkinson, Engineering Conferences International, Tomar, Portugal, July 1 - 6, 2007

DESIGNING TWO-PHASE FLOW HEAT EXCHANGERS FOR MITIGATING FOULING
C.B. Panchal1 and Blazo Ljubicic2
1

E3Tec Service, LLC, 11865 Tall Timber Drive, Clarksville, MD 21029, USA 2 Koch Heat Transfer Company, 12602 FM529, Hoston, TX 77041

Abstract Significant progress has been made in understanding the underlying mechanisms of hydrocarbon fouling and characterizing the effects of physical and chemical parameters in single-phase flows. However, the interactive effects of two-phase flows and fouling mechanisms are poorly understood. Many of the industrial heat exchangers in refining and petrochemical processing operate under multiphase flow conditions with varying flow regimes. Fouling and localized corrosion are generally experienced due to maldistribution of vapor and liquid phases in such heat exchangers. This is because design methods are inadequate to determine localized process conditions that could induce severe fouling locally, which could also lead to localized corrosion by concentrating corrosion species in the deposits. This paper discusses design approaches to evaluate twophase flow conditions and design exchangers to minimize fouling. Such design approaches may not be optimum for heat transfer and pressure drop considerations; however, it will minimize fouling and localized corrosion. The paper focuses on improved design of feed/effluent exchangers with tube side flows and reboilers for minimizing shell-side fouling and a case study is presented. The result is lower lifecycle ownership cost of the exchanger. Introduction Heat exchangers with two-phase flows in petroleum processing that encounter fouling and coking are: Reboilers: localized fouling is generally caused by maldistribution of two-phase flow on the shell side creating localized environment for chemical reaction fouling and particulate deposition of corrosion products which might be

generated in the overhead condenser/separation drums, transfer pipelines, and/or in the distillation column. Feed/Effluent (F/E) exchangers: fouling on tube side of F/E exchangers in hydrotreating and reforming processes is mostly caused by maldistribution of vapor and liquid phases among tubes and intermittent dry out caused by the mist flow at higher vapor qualities. Condensers/coolers: condensers/coolers (including air-cooled condensers) are used downstream of hydrocracking and hydrotreating reactors and they are prone to fouling caused by solid phase formation on cooling. Hydrocarbon process streams in overhead condensers of fractionation columns generally do not produce fouling, but fouling is mostly observed on cooling-water side. Heat recovery boilers: they are commonly used to recover heat from reactor effluent to generate highpressure steam. Other than the transferline exchanger (TLE) in the ethylene process, fouling is not a major problem for heat recovery boilers. Fired process heater: the process side has normally two-phase flows with all possible flow regimes. Fouling/coking is generally localized; normally downstream of U-bends. This study focuses on fouling mechanisms, methodology to determine root cause(s) of fouling in reboilers and F/E exchangers, and guidelines for designing exchangers with twophase flow applications. A case study of mitigating fouling of a reboiler is presented to demonstrate how a proper design approach can mitigate fouling.

Produced by The Berkeley Electronic Press, 2013

org/heatexchanger2007/11 . however. Furthermore. like the idea that higher fouling factors would result in longer cleaning cycles.72 Heat Exchanger Fouling and Cleaning VII [2007]. right? We use certain procedures. Article 11 Reboiler Design Practice The term ‘reboiler’ is generally used in the Process Industry to describe equipment whose function is to add a specified duty to a distillation or other separation process. cannot be reproduced with bench-scale test units. how do we create the evidence the end-users demand. embedded in commercial software suites because they work well. The end users are generally responsible for taking corrective measures to minimize the effects of the two-phase pattern and localized thermal conditions on fouling. Some basic guidelines are: • • Use a proven thermal design tool(s). but it should be effective in dismissing the most cherished beliefs in the industry. Increasingly. don’t we? However. after all is a service built around uniqueness of each particular design and engineers’ ability to design equipment accordingly. In fact. However. empirical look at what works what doesn’t work. particularly at maximum turndown. Vol. only three types of reboilers are common: • • • Kettle. • • • Nobody pretends design of reboilers is straightforward. This current stop-gap practice is generally responsible for fouling of reboilers and costly corrective measures the end users have to implement in after-the-fact scenario. if operational problems are to be avoided the design process of this type of equipment calls for very careful thought. We use certain methods because they work. before releasing design methods to end-users correlations are generally tested with available test data to show that they are effective and reliable. and Horizontal thermosyphon or forced circulation. Often. engineers are left to propose a design based at least in part on engineering judgments and previous experiences. but there is one thing we ought to be able to rely on: it’s that the engineers working on a design should be doing more than playing hunches. Do not specify an excessive fouling resistance. and how to distinguish between the two. RP5. The primary objective of the design is to meet the performance in terms of duty for nonfouled conditions. Check the performance at every possible operating condition. there’s simply little evidence that their predictions are reliable. Separated and mist flow regimes should be avoided. Always check the flow regimes. cold. Vertical thermosyphon or forced circulation. unless we test the untested? Whatever the merits of “evidence based engineering. Consider designing reboilers with low residence time. This approach has been used with varying degree of success. specifically on the shell side. especially at maximum turndown and at start-up (no fouling). this can lead to film boiling when the exchanger is clean and worse performance than if it were fouled.engconfintl. They are broadly classified as once-through. designers seeking to provide end users with the best possible solution are relying on what is known as evidence-based engineering. that no matter how rigorous design methods are. In practice. a hard. more often than not engineers realize http://dc. Obviously. or even an educated guess. which can be achieved with low liquid hold up and minimizing recirculation zones. thermosyphon.” further progress will require an increased reliance on “expert” opinion supported by knowledge-based prediction methods. The need for careful design of reboilers is well known. the two-phase flow with or without heat transfer in industrial heat exchangers. But is this possible or even desirable? Heat transfer engineering. or forced flow.

g. As a result. It was shown that the fouling rate was significantly high in tubes with high liquid to hydrogen flow ratios. Major Fouling Mechanisms. separated flows. 5) distribution of fouling deposits within the exchangers. Therefore. and higher wall temperatures particularly in regions of intermittent vapor-liquid interfaces on the wall. 1999] is a common problem in reboilers and reactor overhead condensers of hydrotreaters. Reaction fouling – polymerization. Amelse et al [2004] analyzed the distribution of vapor and liquid phase in F/E exchanger of p-xylene plant using radioactive tracers. decomposition of organometallic complexes) would occur. the long-term performance of such exchangers under varying process conditions needs to be validated. In practice the following types of F/E exchangers are used. and FCC fractionators. Polymerization fouling. In order to avoid poor vapor and liquid phase distribution. If these localized wall temperature are above threshold fouling conditions at prevailing liquidphase wall shear stress. no corrective measures are built into designing F/E exchangers. However. then formation of precursors (e. generally induced by iron-sulfide formation in refining. there is a need to examine the current practice of designing F/E exchangers. develop validated predictive methods for localized fouling/coking conditions. particularly on how vapor and liquid are fed to the exchanger. All these parameters are responsible for accelerated fouling rates. single phase flow can be applied to two phase flows. Deposition of corrosion products. 2013 . Tubes with low vapor flows would produce slug or separated flows with high residence time and low wall shear stress would increase the precursor formation in the liquid phase. They show low fouling propensity. and evaluate new design configurations including alternate heat exchangers. correlations are not available to determine threshold limits of these parameters for two phase flows in tubes and we are not sure how threshold conditions determined with the Produced by The Berkeley Electronic Press. The root cause analysis should focus on: 1) two-phase flow patterns at the inlet. There are no adequate design guidelines to evaluate fouling propensity for two-phase flow conditions under varying process parameters. • • • Horizontal shell and tube with Utube bundles. Root Cause Analysis and Mitigation The three major fouling mechanisms associated with reboilers and F/E heat exchangers are: 1. 4) thermal profiles. and Vertical welded plate exchangers. generally tube side. vertical welded plate heat exchangers are being considered. Such conditions produce low wall shear in the liquid phase. Tubes with high vapor flows could produce undesired flow regimes (such as mist flows) with intermittent dryout or thin film. Coking under dryout and/or under critical heat-flux conditions. The current design practices focus on achieving the thermal performance for a given pressure drop limit. Polymerization of diolefins [Brons et al. Non-uniform distribution of the petroleum feed and recycled hydrogen among tubes of horizontal as well as vertical F/E exchangers induces high localized wall temperatures. of F/E exchangers of hydrotreating and reforming process units [Limke. is commonly observed on feed side. hydrocrackers.Panchal and Ljubicic: 73 Feed/Effluent (F/E) Exchangers F/E exchangers are used to recover process heat from reactor (hydrotreater and reformer) products and preheat the feed stream. Vertical combined feed exchanger (VCFE) also known as vertical Texas Tower F/E. 3) chemical analysis of deposits collected from different parts of the exchanger to determine the main precursor(s) and hence identify the most probable chemical mechanism. and 3. however. 2) flow regimes in return passes. 2. and 6) change in The two-phase flow feed is normally on tube side and the product stream is on the shell side. which consists of petroleum feed and recycled hydrogen. 1999]. which could produce high fouling/coking. They correlated high rates of fouling with the hydrogen and hydrocarbon distribution.

it is important to evaluate individual applications for suitability of such exchangers.33 exp( -E/RTf) . The root cause analysis should focus on deposit analysis and tracking the source of corrosion. Welded plate heat exchangers with narrow flow passages have been shown to reduce fouling. The rate of fouling is expressed in terms of generation of precursor in the boundary layer and its removal before firm adhesion of deposits on the wall occurs. which can then propagate to the rest of exchangers. Exchangers that provide good distributions of liquid and vapor phases and minimize the overall resident time would minimize fouling by reducing the rate of precursor formation in the exchanger. The localized deposition can alter the flow patterns and thermal conditions locally. where corrosion products may be generated in the distillation column or overhead condensers/separation drums. however. Understanding the governing fouling mechanisms and designing the exchanger to operate below threshold wall conditions (temperature and wall shear stress) provides necessary design information for an alternate tube bundle to minimize fouling. The root-cause analysis should focus on analysis of deposits to determine coking mechanisms (solubility tests) and identifying regions with high coking within the exchanger. producing conditions of higher fouling propensity for organic fouling (polymerization) in other regions. and evaluating threshold coking conditions during turn-down operation (e. Vol. However. Alternate heat exchanger designs which provide good flow patterns on shell side (twisted tubes or spiral baffles) should be considered. Article 11 pressure drop between cleaning. use of recirculation pump) are possible mitigation steps based on design. if mist flow occurs and high-temperature heating media (e.engconfintl. Polymerization fouling may lead to further coking reaction within the deposits as wall temperatures increase after significant fouling occurs. controlling corrosion fouling could minimize organic fouling.g. F/E exchangers with two-phase flow on shell side provide low residence time and should have relatively uniform distributions to minimize fouling. avoiding super heated steam. Maldistribution of liquid and vapor phases at the feed point and also in return passes of multi-pass exchangers can produce conditions with high propensity of localized fouling. If alternate tube bundle cannot meet necessary conditions to minimize conditions then consider altogether different heat exchanger. Such fouling data or an empirical correlation based on these data cannot be directly applied to exchangers with two-phase flows. The reboiler in general should have provision purge corrosion products collected at the feed point.g. Panchal et al [1997] developed a threshold fouling model applicable to preheat train. The top region of reboilers is prone to coking. RP5. Intermittent wetting and dryout produce droplets and thin film and the transient heat transfer induces coke-like deposits separating out as solid or semi-solid phase. Application of Single-Phase Fouling Data to Two-Phase Flows Fouling data are generally obtained with singlephase flow fouling units. The deposition of corrosion products is more common in reboilers.λ τs (1) http://dc. These corrosion products tend to deposit in recirculation zones. then Kuru and Panchal [1997] showed how the fouling data obtained with single-phase fouling unit could be applied to tube side annular twophase flow regimes. twisted tubes. and spiral baffles to alter thermal profiles and two-phase flow patterns. The primary choices for alternate tube bundles are different tube size.74 Heat Exchanger Fouling and Cleaning VII [2007]. highpressure steam) is used. dRf/dt= α Re -β Pr -0. Improved or alternate heat exchanger design to minimize mist flow. More commonly coking in reboilers and F/E exchangers is caused by dryout and critical heat flux with high wall temperatures. If the flow regimes are known and reliable correlations are available to predict the local wall shear stress and heat transfer rates. Combined corrosion and polymerization fouling can be difficult to manage.org/heatexchanger2007/11 .

respectively. Similar flow distribution is expected for horizontal F/E exchanges with U-bend tube bundles.3 0. which is commonly used by the petroleum industry to estimate TMTs.0 0. The distribution of vapor and liquid phases in the U-bend regions further increases propensity of localized coking.1 0.8 1.6 0. Constants α. The wall shear stress. Figure 1 shows circumferentially averaged tube-metal temperature (TMT) for selected pipes of a crude heater predicted by the process model and API 530 (American Petroleum Institute design standard 530).2 0 0. The model was developed for single-phase flows.8 0. it provides threshold fouling/coking conditions between in-side tube metal temperature and the wall shear stress. β. Re and Prandtl number. Panchal et al.Panchal and Ljubicic: 75 Where Reynolds number.6 0. however.4 0.5 0. it can be applicable to two phase flows provided local fluid dynamics can be calculated.7 0.0 0. The predicted results show gradual increase in the coking rates as the TMT temperature rises. 1 developed for crude preheat train to the process side of a crude fired heater. 840 820 800 780 760 TMT. the heat fluxes and hence TMTs are incorporated.1 0. The Reynolds number in this equation was based on the two-phase flow viscosity.4 Coking rate 1. λ and the activation energy E are determined from laboratory and/or field data.9 1.6 1. Pr are calculated for properties at the bulk and film temperature Tf (average of bulk and metal wall temperatures). the localized coking rates increases significantly.5 0. This case study is presented here to demonstrate applicability of the fouling model to an industrial fired heater to evaluate the localized fouling propensity.2 1 0. 2 1.6 0.0 Axial distance from top Pipe#1 Pipe#13 Pipe#2 Pipe#14 Pipe#29 Pipe#25 Pipe#30 Pipe#26 Fig. [2006] applied the fouling correlation in Eqn.3 0.0 Axial distance from top Pipe#1 Process Model Pipe#30 Process Model Pipe#1 API530 Pipe#30 API530 Fig. τs is calculated for the liquid flow at the wall surface. 3 Cross section at up flow exit of a down flow U-bend Figure 3 shows secondary flow patterns and accumulation of liquid (dark region) on the outer Produced by The Berkeley Electronic Press. F 740 720 700 680 660 640 620 600 0.2 0. although the wall shear stress also increases with increase of the vapor quality.7 0.2 0.4 0.8 0.4 0.9 1. 1 TMTs predicted by the process model and API-530 The localized coking rate predicted using Eqn.2 Coking rate (hr ft2 F/Btu (10-4) per day) distribution. 1 is shown in Figure 2. When the equation is equated to zero. 2013 . When the circumferential variations of Fig. Figure 3 shows the CFD-based prediction of distribution of vapor and liquid phase downstream of U-bend.8 0.

Using the process data to make direct comparisons about fouling in different geometries is difficult in most cases. http://dc. temperature. [1988] showed that the heat transfer coefficient of flow boiling in vertical spirally fluted tubes remains relatively high for vapor qualities 90% to 95%. Panchal et al.76 Heat Exchanger Fouling and Cleaning VII [2007]. velocities. the condition at which vapor blanketing is initiated causing accelerated fouling. the swirl flow (with its secondary flows in the boundary layer) is expected to enhance the fouling removal rate from the surface. This will significantly improve design capabilities of exchangers with two-phase flows and minimize fouling. In fact. and concentration. Figure 4. Therefore. twisted tube exchangers. In addition. concentrations. it should be noted that Fig. nucleate boiling characteristics of twisted tubes are similar to those of conventional. reaction rates are quite different from those encountered in conventional rounded tubes. temperatures. There are several reasons for this. However. Article 11 wall of the bend and liquid distribution at midplane of a down flow U-bend. for the same pressure drop. rounded tubes. rounded tubes. However. very little work has been done to develop such predictive methods for fouling in two-phase flows. This case study shows potential application of a fouling model developed and validated with single-phase flow data applied to two-phase flows.engconfintl. lower wall temperatures and thereby minimize the fouling propensity. shown as dotted line. Thus. offer unique advantages in F/E exchanger applications. Within the boundary layers of twisted tube flows. Vol. compared to plain tube units. the critical flux of twisted tubes (inside and outside) is approximately twice that of straight. The resulting effects of minimizing the intermittent (slug and plug) and mist flows maintains good distribution of liquid phase on the wall surface.org/heatexchanger2007/11 .Twisted Tube Technology Swirl flow velocity and temperature field generated by twisted tubes. the future research and development should focus on developing and validating the fouling prediction correlation for different flow regimes. Studies show that boiling inside and outside twisted tubes is effective in increasing critical heat fluxes. as a result of momentum and heat and mass transport phenomena occurring within the boundary layer. In comparing twisted-tube bundle to conventional baffled heat exchangers. tend by its very nature to have lower tendency to fouling Swirling flows in twisted and spirally fluted tubes alter the flow pattern and extend the annular flow regime. RP5. all surface-induced reactions as well as heat-induced reactions take place within the boundary layers of velocity. all develop at the heating surfaces. Figure 5 shows that twisted tube dramatically shifts the mist flow regime.4 Tube-side/shell-side swirl flow enhancement in TT bundles twisted tube bundles have a more uniform flow and temperature field with almost no shell-side dead areas providing better conversion of available pressure drop to heat transfer. First. which maintain high heat transfer coefficient. as compared to vertical circular tube where the performance drops significantly for vapor qualities greater than 80%. Swirl Flow .

6 Convective boiling in twisted tubes The thermal performance results clearly indicate that tubes with swirling flows should be considered for F/E exchangers with high vapor qualities to minimize fouling by maintaining extended annular flows and minimizing the high localized wall temperatures. including kettles. indicating potential longer runtime. the performance would drop for lower vapor qualities.Panchal and Ljubicic: 77 required to instigate nucleate boiling could be high. With fresh liquid entering the bundle from five directions. However. Kettles are not pool boiling devices and their operation is limited by the flow boiling critical flux. The objective of the installation was to extend the mean runtime between reboiler cleaning and debottleneck depropaniser column capacity. data from installation of a Twisted Tube bundle in a reboiler service is discussed. The horizontal twisted tubes maintain relatively high heat transfer coefficient for vapor qualities upto about 80% as shown in Figure 6. One option is the use of special boiling surfaces. Twisted Tube bundle geometry (Figure 7) provides better recirculation rate compared with only three in the conventional configuration. As we did not have all operating data required to calculate heat transfer coefficients on a daily basis. Fig. 5 Comparison of two-phase flow regimes in rounded and twisted tubes In horizontal tubes. An analysis of initial startup conditions quickly revealed that Twisted Tube bundle was operating at considerably lower steam pressure. 2013 . Here. 7 Twisted Tube geometry This limit is not related to the boiling crises observed in pool boiling experiments. bundle. these are expensive and investments hard to installations. Case Study The future improvements in both increased energy efficiency and extended runtimes will likely come from the use of closer temperature approaches. Revamping with Twisted Tube heat exchangers is often the cheapest and most cost effective way to debottleneck the plant. Data available in the open literature indicate that the most probable cause of performance degradation in conventional kettle reboilers is due to insufficient liquid recirculation through the Fig. conventional exchangers will in most cases pose a severe limitation. reboiler performance and fouling was assessed or ‘trended’ using indirect process Fig. As the temperature difference Produced by The Berkeley Electronic Press.

160 psig for the steam condensate drum. In the short term. However. design methods have progressed significantly.8 Comparison of DeC3 reboilers Conclusions An overview of technical issues of designing heat exchangers with two-phase flows is presented and how significant fouling/coking can occur due to poorly designed reboilers and F/E exchangers. Article 11 parameters such as opening of the flow control valve (FCV) steam or the pressure of the reboiler steam condensate drum. Since there is an upper limit on condensate pressure (steam pressure is approximately 160 psig). Alternate designs are available to reduce two-phase flow maldistribution and thereby minimizing conditions for high fouling and coking. FCV opening and condensate pressure provided reliable fouling rate trends.78 Heat Exchanger Fouling and Cleaning VII [2007]. Figure 8 provides a comparison of the fouling rates for five reboilers. Two-phase fouling beyond the critical conditions. In this analysis. pressure readings and valve openings are significantly more accurate with fewer variations and therefore easier to analyze. film boiling). indicating more available surface area and higher cleaning efficiency. Vol. Fouling in kettle reboilers is common and unfortunately not well understood. http://dc. It is important to notice that the start of run pressure is significantly lower in some reboilers. the interactive effects of two-phase flows and fouling mechanisms cannot be predicted. Fig. The major technical issues of designing two-phase flow exchangers and minimize fouling the propensity are as follows: Although. The trend line of steam pressure is a linear correlation of the operating data depicting the fouling rate. the fact that plant is operating three parallel fractionation trains further helped in the analysis of reboiler fouling. the governing fouling mechanisms should be identified and the effects of process parameters on fouling should be evaluated when designing or revamping existing reboilers and F/E exchangers. it is encouraging to see that the same mechanisms that enhance thermal performance also reduce fouling propensity. however. and the Twisted Tube reboiler. Understanding of many fouling mechanisms is incomplete.org/heatexchanger2007/11 . RP5. this application proved that fouling rate within Twisted Tube bundles has been significantly reduced. however. In order to take corrective steps at the design level. can be improved by use of Twisted Tube technology. Although it remains unclear how to cast the observed improvements into quantitative correlations.engconfintl. low condensate pressure at startup should translate into longer runtimes. 100% for the FCV and 155 . two decontamination cleaned. feed composition and light key bottom specification can affect the results. In addition. over an extended period. both the opening of the steam FCV and the pressure of the steam condensate drum had a definite upper limits. Potentials of making design mistakes for reboilers and F/E exchangers are greater than that for designing single-phase exchanger under comparable fouling propensity. The figure shows that the Twisted Tube has the lowest rate of fouling (slope of steam pressure vs time). (mist flow. two that were acid cleaned. Other parameters such as steam rate to each reboiler were also evaluated. and the ability to predict fouling rates is limited. operating variable such as column feed rates.

“Fouling in Refinery Equipment – an Overview. pp. “Mechanisms of Reboiler Fouling During Reforming. Norwood SD. and Stellman RS. Brons G and Wiehe IA. Vol. New York. As energy and efficiency of thermal systems become more important factors in the overall economics of process industry. Kuru WC. 2. and Palen JW. NY. 2013 . New York. “Threshold conditions for crude-oil fouling. the broader acceptance of enhancement technologies in industrial practice will take place.” Proceedings of Heat Transfer Equipment in the Process Equipment. and Mixon W. Kuru WC and Panchal CB. TX. TX. 273-281. Produced by The Berkeley Electronic Press. AIChE Spring Meeting. Begell House Publication. Inc. Panchal CB. There is limited amount of data on critical flux with geometries and conditions relevant large industrial designs. Panchal CB. on Multi-Phase Transport and Particulate Phenomena. 2004.” Proceedings of the AIChE National Meeting. 1997. and Petrick M. “Two-phase flow pipe-side process model for fired heater. LA.Panchal and Ljubicic: 79 The uncertainty of the appropriate value for the critical flux and onset of fouling for horizontal F/E exchangers grows as the size of the unit becomes larger. References Amelse JA. Liao CF. editor. 1997. AIChE Spring Meeting. Publisher Begell House. Houston. "Forced-Convective Vaporization of Refrigerants in Spirally Fluted Tubes. Symp. Orlando. 2006. FL. Panchal CB. France DM. TR Bott..” Proceedings of the Engineering Foundation Conference: Understanding Heat Exchanger Fouling and Its Mitigation. 43. 1988. of the Intl. Limke H 1999. NY." Proc.” Proceedings of the AIChE National Meeting. “A radioactive tracer study of gas and liquid distribution in horizontal CEN paraxylene unit feed/effluent heat exchangers. Ebert WA. New Orleans. “Application of single-phase flow fouling date to two-phase flows. Houston.” Proceedings of Heat Transfer Equipment in the Process Equipment. Lottes S. 1999.” Understanding Heat Exchanger Fouling and Its Mitigation.

Sign up to vote on this title
UsefulNot useful